Published July 17, 2026
| Version 0.3
Dataset
Open
RUBATO: A Multi-Version Benchmark for Robust Music Transcription and Analysis
Authors/Creators
Description
01_Raw_Data
- score_image: PNG image exports from the symbolic score
- score_midi: Midi exports from the symblic score
- score_musescore: Symbolic score in MuseScore format
- score_pdf: Sybolic score in PDF format
- video_30fps: Video recording of the piano performance corresponding to the MIDI file of the score
- wav_22050_mono: Audio file
02_Annotations
- ann_audio_beat: Beat annotations given in physical time
- ann_audio_lokalkey: Local Key annotations given in physical time
- ann_audio_measure: Measure annotations given in physical time
- ann_audio_noteEvents: Note events with start, end, velocity and instrument given in physical time
- ann_audio_startEnd: Start and end time of tracks
- ann_audio_structure: Structure annotations given in phyiscal time
- ann_score_localkey: Local key annotations given in musical time
- ann_score_structure: Structure annotations given in musical time
- scripts: Python scripts for preprocessing and allignment
03_ExtraMaterial
- scripts: Python scripts for preprocessing and allignment
- warpingPaths: Warping paths aligning each audio track to corresponding score